5 files changed, 111 insertions, 83 deletions

diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 49b7a4c31a6d..2483a05231c7 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -34,83 +34,10 @@ struct sg_io_hdr;
 #define BLKDEV_MIN_RQ   4
 #define BLKDEV_MAX_RQ   128     /* Default maximum */
 
-/*
- * This is the per-process anticipatory I/O scheduler state.
- */
-struct as_io_context {
-        spinlock_t lock;
-
-        void (*dtor)(struct as_io_context *aic); /* destructor */
-        void (*exit)(struct as_io_context *aic); /* called on task exit */
-
-        unsigned long state;
-        atomic_t nr_queued; /* queued reads & sync writes */
-        atomic_t nr_dispatched; /* number of requests gone to the drivers */
-
-        /* IO History tracking */
-        /* Thinktime */
-        unsigned long last_end_request;
-        unsigned long ttime_total;
-        unsigned long ttime_samples;
-        unsigned long ttime_mean;
-        /* Layout pattern */
-        unsigned int seek_samples;
-        sector_t last_request_pos;
-        u64 seek_total;
-        sector_t seek_mean;
-};
-
-struct cfq_queue;
-struct cfq_io_context {
-        struct rb_node rb_node;
-        void *key;
-
-        struct cfq_queue *cfqq[2];
-
-        struct io_context *ioc;
-
-        unsigned long last_end_request;
-        sector_t last_request_pos;
-
-        unsigned long ttime_total;
-        unsigned long ttime_samples;
-        unsigned long ttime_mean;
-
-        unsigned int seek_samples;
-        u64 seek_total;
-        sector_t seek_mean;
-
-        struct list_head queue_list;
-
-        void (*dtor)(struct io_context *); /* destructor */
-        void (*exit)(struct io_context *); /* called on task exit */
-};
-
-/*
- * This is the per-process I/O subsystem state.  It is refcounted and
- * kmalloc'ed. Currently all fields are modified in process io context
- * (apart from the atomic refcount), so require no locking.
- */
-struct io_context {
-        atomic_t refcount;
-        struct task_struct *task;
-
-        unsigned int ioprio_changed;
-
-        /*
-         * For request batching
-         */
-        unsigned long last_waited; /* Time last woken after wait for request */
-        int nr_batch_requests;     /* Number of requests left in the batch */
-
-        struct as_io_context *aic;
-        struct rb_root cic_root;
-        void *ioc_data;
-};
-
-void put_io_context(struct io_context *ioc);
+int put_io_context(struct io_context *ioc);
 void exit_io_context(void);
 struct io_context *get_io_context(gfp_t gfp_flags, int node);
+struct io_context *alloc_io_context(gfp_t gfp_flags, int node);
 void copy_io_context(struct io_context **pdst, struct io_context **psrc);
 void swap_io_context(struct io_context **ioc1, struct io_context **ioc2);
 
@@ -894,6 +821,12 @@ static inline void exit_io_context(void)
 {
 }
 
+static inline int put_io_context(struct io_context *ioc)
+{
+        return 1;
+}
+
+
 #endif /* CONFIG_BLOCK */
 
 #endif
diff --git a/include/linux/init_task.h b/include/linux/init_task.h
index 796019b22b6f..e6b3f7080679 100644
--- a/include/linux/init_task.h
+++ b/include/linux/init_task.h
@@ -137,7 +137,6 @@ extern struct group_info init_groups;
                 .time_slice     = HZ,                                   \
                 .nr_cpus_allowed = NR_CPUS,                             \
         },                                                              \
-        .ioprio         = 0,                                            \
         .tasks          = LIST_HEAD_INIT(tsk.tasks),                    \
         .ptrace_children= LIST_HEAD_INIT(tsk.ptrace_children),          \
         .ptrace_list    = LIST_HEAD_INIT(tsk.ptrace_list),              \
diff --git a/include/linux/iocontext.h b/include/linux/iocontext.h
new file mode 100644
index 000000000000..593b222d9dcc
--- /dev/null
+++ b/include/linux/iocontext.h
@@ -0,0 +1,95 @@
+#ifndef IOCONTEXT_H
+#define IOCONTEXT_H
+
+#include <linux/radix-tree.h>
+
+/*
+ * This is the per-process anticipatory I/O scheduler state.
+ */
+struct as_io_context {
+        spinlock_t lock;
+
+        void (*dtor)(struct as_io_context *aic); /* destructor */
+        void (*exit)(struct as_io_context *aic); /* called on task exit */
+
+        unsigned long state;
+        atomic_t nr_queued; /* queued reads & sync writes */
+        atomic_t nr_dispatched; /* number of requests gone to the drivers */
+
+        /* IO History tracking */
+        /* Thinktime */
+        unsigned long last_end_request;
+        unsigned long ttime_total;
+        unsigned long ttime_samples;
+        unsigned long ttime_mean;
+        /* Layout pattern */
+        unsigned int seek_samples;
+        sector_t last_request_pos;
+        u64 seek_total;
+        sector_t seek_mean;
+};
+
+struct cfq_queue;
+struct cfq_io_context {
+        void *key;
+        unsigned long dead_key;
+
+        struct cfq_queue *cfqq[2];
+
+        struct io_context *ioc;
+
+        unsigned long last_end_request;
+        sector_t last_request_pos;
+
+        unsigned long ttime_total;
+        unsigned long ttime_samples;
+        unsigned long ttime_mean;
+
+        unsigned int seek_samples;
+        u64 seek_total;
+        sector_t seek_mean;
+
+        struct list_head queue_list;
+
+        void (*dtor)(struct io_context *); /* destructor */
+        void (*exit)(struct io_context *); /* called on task exit */
+};
+
+/*
+ * I/O subsystem state of the associated processes.  It is refcounted
+ * and kmalloc'ed. These could be shared between processes.
+ */
+struct io_context {
+        atomic_t refcount;
+        atomic_t nr_tasks;
+
+        /* all the fields below are protected by this lock */
+        spinlock_t lock;
+
+        unsigned short ioprio;
+        unsigned short ioprio_changed;
+
+        /*
+         * For request batching
+         */
+        unsigned long last_waited; /* Time last woken after wait for request */
+        int nr_batch_requests;     /* Number of requests left in the batch */
+
+        struct as_io_context *aic;
+        struct radix_tree_root radix_root;
+        void *ioc_data;
+};
+
+static inline struct io_context *ioc_task_link(struct io_context *ioc)
+{
+        /*
+         * if ref count is zero, don't allow sharing (ioc is going away, it's
+         * a race).
+         */
+        if (ioc && atomic_inc_not_zero(&ioc->refcount))
+                return ioc;
+
+        return NULL;
+}
+
+#endif
diff --git a/include/linux/ioprio.h b/include/linux/ioprio.h
index baf29387cab4..2a3bb1bb7433 100644
--- a/include/linux/ioprio.h
+++ b/include/linux/ioprio.h
@@ -2,6 +2,7 @@
 #define IOPRIO_H
 
 #include <linux/sched.h>
+#include <linux/iocontext.h>
 
 /*
  * Gives us 8 prio classes with 13-bits of data for each class
@@ -45,18 +46,18 @@ enum {
  * the cpu scheduler nice value to an io priority
  */
 #define IOPRIO_NORM     (4)
-static inline int task_ioprio(struct task_struct *task)
+static inline int task_ioprio(struct io_context *ioc)
 {
-        if (ioprio_valid(task->ioprio))
-                return IOPRIO_PRIO_DATA(task->ioprio);
+        if (ioprio_valid(ioc->ioprio))
+                return IOPRIO_PRIO_DATA(ioc->ioprio);
 
         return IOPRIO_NORM;
 }
 
-static inline int task_ioprio_class(struct task_struct *task)
+static inline int task_ioprio_class(struct io_context *ioc)
 {
-        if (ioprio_valid(task->ioprio))
-                return IOPRIO_PRIO_CLASS(task->ioprio);
+        if (ioprio_valid(ioc->ioprio))
+                return IOPRIO_PRIO_CLASS(ioc->ioprio);
 
         return IOPRIO_CLASS_BE;
 }
diff --git a/include/linux/sched.h b/include/linux/sched.h
index df5b24ee80b3..2d0546e884ea 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -27,6 +27,7 @@
 #define CLONE_NEWUSER           0x10000000      /* New user namespace */
 #define CLONE_NEWPID            0x20000000      /* New pid namespace */
 #define CLONE_NEWNET            0x40000000      /* New network namespace */
+#define CLONE_IO                0x80000000      /* Clone io context */
 
 /*
  * Scheduling policies
@@ -975,7 +976,6 @@ struct task_struct {
         struct hlist_head preempt_notifiers;
 #endif
 
-        unsigned short ioprio;
         /*
          * fpu_counter contains the number of consecutive context switches
          * that the FPU is used. If this is over a threshold, the lazy fpu